Improved geometric filter algorithm for device free localization

Device Free Localization (DFL) is one of the emerging applications of Wireless Sensor Networks, which can find people or objects in a monitored area — even through walls. ‘Device Free’ — which states that the method can localize a person or target, without the target using any Passive or active devices such as RFID tags or GPS devices. ‘Tracking through walls’ — which make the method unique than other device free monitoring methods like the ones employing Image/Video processing or with Infrared or Light rays. This method uses Radio signals to scan an area to localize the target. The Geometric Filter (GF) algorithm discussed in this paper is one of the approaches which uses Received Signal Strength (RSS) in the links as a metric and gives better tracking performance using only simple computations based on geometric objects. This method performs better than others with lower tracking errors and with huge improvement in average execution time. This paper analyses the Geometric Filter algorithm with performance evaluation in different number of nodes and with measurement noise. An improvement is also suggested.